Why Huawei Cloud GeminiDB?
Shared-nothing and shared storage architecture: Stateless compute nodes share a stateful distributed storage engine.
High resource utilization: Built-in proxy clusters support dynamic data load balancing.
Elastic scaling: Compute nodes can be added in minutes and storage can be scaled up in seconds.
No data loss & Strong data consistency: Three copies of data are stored in a distributed shared storage pool, ensuring zero data loss and strong consistency for multi-point access.
High availability: Superlative fault tolerance (N-1 reliability) ensures that if a node is faulty, services can be taken over in seconds with little impact on services.
Stable performance: GeminiDB has stability designed in, from end to end, with distributed multithreading at the compute layer and remote direct memory access (RDMA) at the storage layer. The RocksDB storage engine is optimized, and cold and hot data separation accelerates hotspot data access.
Dedicated compute: With decoupled storage and compute, GeminiDB leverages dedicated compute nodes to boost computing performance and improve the compute resource utilization of standby servers that use traditional networks.
Low memory cost: GeminiDB Redis API has less demanding memory requirements than open-source Redis, which reduces hardware costs by as much as 75% to 90%.
Low storage cost: For time series data, with the support for adaptive compression and automated hot and cold tiered storage, GeminiDB Influx API requires you to pay only 1/10 of the storage cost of a relational database.
Key-value data and traditional Redis data structures like hashes, lists, sorted sets, sets, and streams
Ads: To store feature profiles, AI services require low-latency and key-value (KV) databases that can hold lots of data. Media terminals also require low latency. GeminiDB Redis API is a great fit for advertising services using real-time advertising (RTA) and on the application platform, demand-side platform (DSP).
Gaming: GeminiDB Redis API helps game companies store player data, activities, rankings, friends, and more. It helps them respond quickly to data issues with Point-In-Time Recovery (PITR).
E-commerce: GeminiDB Redis API enables users to store massive volumes of inventory product data and provides excellent reliability.
GeminiDB Redis API is compatible with Redis APIs and supports primary/standby and cluster architectures.
GeminiDB Redis API is compatible with DynamoDB APIs.
Document type data, which is saved in a JSON-like format
Boolean, numeric, string, date and time, array, object, and nested composite types
Social networking: A huge amount of unstructured data is generated on social networking applications. GeminiDB MongoDB instances are great for storing a large amount of user information, such as user moments, and location data.
IoT: GeminiDB Mongo API can handle high volumes of concurrent requests, for example, logs reported by smart devices.
E-commerce: Order details, order status, and logistics information are stored in embedded arrays and continuously updated. All order changes can be read in one query. GeminiDB Mongo API is applicable to e-commerce orders, favorites, and shopping cart scenarios.
Gaming: Data such as user information, player items, and bonus points is stored in embedded documents for fast query and update.
GeminiDB Mongo API is compatible with MongoDB APIs and supports replica sets.
Wide table model, column fields with scalability, and structured and semi-structured data
Petabytes of data
Social networking: GeminiDB Cassandra API can store user data such as their followers, posts, and likes, and it can handle content distribution data for social networking applications. It can smoothly scale up to handle traffic surges when there are hot topics trending on applications such as Facebook, Discord, and MeeTime.
IoT: GeminiDB Cassandra API handles concurrent writes of massive data generated by IoT devices, including data about logistics, wearables, and smart city monitoring.
Data warehouse: GeminiDB Cassandra API can re-analyze massive data about advertisements, historical orders, feeds, messages, logs, and recommendations to form new user and behavior profiles.
GeminiDB Cassandra API is compatible with open-source Cassandra, with a development experience similar to that of MySQL and supporting cluster architectures.
GeminiDB Cassandra API is compatible with open-source HBase (in OBT).
GeminiDB Cassandra API is compatible with DynamoDB APIs and data can be imperceptibly migrated from DynamoDB to GeminiDB Cassandra API.
Time series data with timestamps
Simplified data types including boolean, int, string, and float
A large volume of data which is mainly written and queried and seldom updated
Stable data traffic without obvious fluctuations between peak and off-peak hours
Structured data and data generated from a single data source
Data which is stored by time and deleted upon expiration
IoV: information such as vehicle status, location, and battery status
IoT: reported status data such as factory devices, terminals, and environments
Application O&M and monitoring: CPU, memory, and disk usage of VMs and containers
Business intelligence: aggregation calculating and analysis of data such as service records and user behaviors
GeminiDB Influx API is compatible with InfluxDB APIs and supports cluster architectures.